Jacob Roach / Digital Trends
Nvidia has announced its new line of GPUs , the RTX 50 - series — and the first two are almost here , quick to touch thebest graphics cards . We were already able to get our hand on the RTX 5090 , which is why we now have a good estimate of what these menu are able of .
While we ’re still waiting for the RTX 5080 , RTX 5070 Ti , and RTX 5070 , we know that Nvidia is promising some huge leaping in performance , thanks to the new AI powers ofDLSS 4 . Here ’s everything you ask to do it aboutNvidia ’s RTX 50 - serial publication .
Jacob Roach / Digital Trends
RTX 50-series: pricing and release date
To say there was guess about the cost of the RTX 50 - serial is a serious understatement . Well , the leaks and rumors can finally fare to an final stage . Here are the price of forthcoming GPUs :
That ’s a $ 400 increase on the top remainder , up from the $ 1,599 for theRTX 4090 , which was already the most expensive Nvidia GPU ever made . That toll jump is definitely an brow - raiser , but further down the stack , the pricing in reality is n’t defective .
The RTX 5080 matches the price of theRTX 4080(orRTX 4080 Super ) , for object lesson . Meanwhile , the RTX 5070 Ti and RTX 5070 are in reality $ 50 cheaper than theRTX 4070 TiandRTX 4070 . How ’s that for a surprise ?
These just don’t reflect reality.Nvidia
Of of course , cards made by Nvidia ’s partner may push these prices to young high . They ’re not in stock yet , so we ca n’t confirm what we ’ll get , but some early leaks show the RTX 5090 being sell for as much as $ 2,700 . We ’ll have to wait and see the prices once the bill of fare are in stock .
The RTX 5090 and 5080 both set in motion on January 30 , 2025 . The RTX 5070 Ti and 5070 , meanwhile , followed in February and March .
RTX 50-series: specs
There ’s a lot to take in here . We also have it off that the cards will all use a 16 - pin connector and DisplayPort 2.1 , which is where the support for 8 KB resolutions at 120Hz comes from . The placard are also using the 12V-2×6 power connection . This can almost be considered overkill for most of the lineup , but the 575 - watt RTX 5090 may not have a band of overclocking headway with a individual 600 - James Watt power connexion .
luckily , Nvidia claim thatmelting power connectorswon’t be coming back . The 12VHPWR connector that used to dethaw on the RTX 4090 has been replaced with the 12V-2×6 , which , harmonise to Nvidia , will be a safer option .
While we have the official eyeglasses , there ’s a few more details that we likely wo n’t have affirm until reviewer get their hands on the cards .
Image used with permission by copyright holder
The initial announcement did n’t deliver some key information , including retention speed and unconscious process thickening , which Nvidia has since render . As a upshot , we know that the RTX 5080 will have debauched retentiveness faculty than any of the other GPUs . This has long been a subject of speculation , and has now been confirmed by Nvidia .
Overall , we ’re looking at gains all across the board here in many specs , but not in all of them . Unsurprisingly , the RTX 5090 seduce the grown wins , with a huge memory addition ( to a whopping 32 GB , no less ) and a rise to CUDA cores . The other cards ca n’t enjoy the same kind of treatment , with all three keep on the same VRAM and motorcoach configuration . However , the totality of the RTX 50 - series uses much immobile GDDR7 memory , which is an rising slope in itself .
RTX 50-series: mobile GPUs
Somewhat unexpectedly , Nvidia ’s keynote was n’t solely focused on desktop GPUs . In fact , play laptopsplayed a large part in the presentation , and there ’s a pot for notebook gamers to get excited about .
Nvidia and its partners are readying up laptop with almost the full scope of RTX 50 - series wag , include the RTX 5090 , RTX 5080 , RTX 5070 Ti , and the RTX 5070 . The RTX 5060 Ti , RTX 5060 , and the RTX 5050 have n’t been mention yet , but in all likelihood , some ( if not all ) will be coming at a later date .
The only thing we know about these young laptop GPUs is the AI tera operations per 2d ( TOPS ) count for each , start with 1,850 TOPS for the RTX 5090 ( which is comparable to the screen background RTX 5080 ) and terminate with 800 AI top for the RTX 5070 . That laptop computer is meant to be mighty telling , though , as Huang claims that it ’ll touch the performance of an RTX 4090 . Desktop or laptop computer version ? It ’s unclear right now , but I ’m approximate laptop .
Image used with permission by copyright holder
Nvidia has also shared the starting prices for each laptop . The RTX 5070 models start at $ 1,299 , follow by $ 1,599 and up for the RTX 5070 Ti and $ 2,199 for the RTX 5080 . Laptops equipped with the RTX 5090 get really expensive , with the cheapest models priced at $ 2,899 .
Many of these laptops will become available in March . Nvidia ’s collaborator , such asMSI , have already beleaguer some of their upcoming offerings .
RTX 50-series: design
The Founders Edition RTX 5090 might be the bragging surprise with its 2 - slot plan . In fact , the RTX 5090 and 5080 are the accurate same size of it , much like the bigger RTX 4080 and 4090 . The new blueprint still uses two fans , but they ’re now on the same side , unlike the previous generation , which used a dual axial flow rate design . check that to check outour unboxing of the RTX 5090to learn more about Nvidia ’s latest aura card .
The RTX 5090 sport a sensational design , and the fact that it ’s a dual - slot GPU means that it looks shockingly flimsy when liken to the RTX 4090 . The one downside we ’ve found is that the 16 - peg power connector is angled , which can create an takings in some builds , peculiarly if you ’re extend a small manikin factor build .
The rest of the GPUs are n’t out yet , but I expect that they ’ll be like design - judicious in their Founders Edition variants . Like in anterior years , there is no Founders Edition of the RTX 5070 Ti .
Nvidia
Of of course , control panel married person will no doubt make much with child versions of these wag that take up more than two one-armed bandit . In fact , so far , the Founders Edition of the RTX 5090 is the only one that ’s ready for Nvidia ’s SFF program , incriminate that better half cards are going to be importantly with child .
RTX 50-series: architecture
We know for a fact that the follow - up to Ada Lovelace is called Blackwell , honoring American mathematician David Blackwell . The new consumer GPUs are manufactured based on TSMC ’s 4NP mental process , which is a 4 nm node . other leak about the card suggested that Nvidia would shift to a 3 Land of Enchantment operation , but that did n’t pan out .
The batting order includes chip starting from the high - ending GB202 through the GB203 , GB205 , GB206 , and entry - level GB207 . We have n’t view the GB207 yet , though . The RTX 5070 runs on the GB205 architecture , so presumably , the RTX 5060 will utilise GB207 , but Nvidia has yet to herald the budget part of the lineup .
Nvidia ’s Blackwell architecture brings a lot of improvements . Unsurprisingly , the RTX 5090 is Nvidia ’s crown accomplishment , and that ’s the one GPU that let a lot of inscrutable dive . As an exemplar , Nvidia spoke to Digital Foundry about the architectural betterment in Blackwell .
Jacob Roach / Digital Trends
The tech heavyweight had to totally redefine the printed circuit board ( PCB ) to support the kind of power consumption ( and just overall exponent , really ) that the new architecture can supply . The GPU that Nvidia blab out about was , of course of action , the RTX 5090 with the GB202 microchip .
The PCB is split into three parts , with the master PCB equip with the massive 30 - form VRM , the plenteous total of VRAM , and the GB202 chip . This is the only part of the PCB that Nvidia unveil during the intro , which — although pocket-size — still houses a brawny 92 - billion - electronic transistor GPU . The missing pieces include video display production , the PCIe storage-battery grid , and office connector for fans . It ’s truly a design we have never see before .
consider that even the monstrous RTX 5090 is just a double - slot card , Nvidia had to do more than just implement a new PCB design . It also gave the cooling organisation a moment of an inspection and repair , using a fluent metallic element thermic port material ( TIM ) instead of thermal paste .
Image used with permission by copyright holder
Nvidia also stick in RTX Neural Shaders that bring AI to programmable shaders . This , in turn , will boost ignition and other quality - colligate feature in games , all in real - fourth dimension .
Although Blackwell is still so new , it ’s already seen its fair portion of troubles . A recent reputation from The Information say us thatNvidia ’s data center GPUs are overheating , which has resulted in delays in the deployment of server rack used for AI preparation . Let ’s trust that with the wildly different pattern factor and the considerate approach to cool down , these problem will never make it to consumer GPUs .
RTX 50-series: DLSS 4
Jensen Huang says thatNvidia ’s Deep Learning Super Sampling 4can “ see into the futurity . ” It ’s a bold claim , but based on other selective information , the update upscaling and frame genesis technical school sound like it ’ll take on a large role in the achiever of this new generation of GPUs .
DLSS 4 , and its neuronal rendering capabilities , is how Huang can explain the fact that the $ 550 RTX 5070 can reportedly rival the $ 1,600 RTX 4090 . During the tonic , Huang explicate that most of the pixels are generated using Nvidia ’s Tensor core , and as a result , only the want picture element are render , while the majority is return with AI .
“ The future of estimator graphics is neuronic rendering , ” Huang suppose .
Image used with permission by copyright holder
This is suppose to supercharge frames per second ( fps ) by up to eight time , and Nvidia ’s example benchmarks show that . The RTX 5090 is tell to be capable to hit an impressive 240 Federal Protective Service at 4 K with full ray trace , which presumably also include course trace .
Nvidia first introduced shape coevals with DLSS 3 , but DLSS 4 come with a new , importantly faster AI model that cuts back on the computational cost to deliver better operation . The modeling is said to be up to 40 % quicker while also being up to 30 % less VRAM - intensive . Nvidia also got rid of the hardware - based optic flow we ’ve seen in the RTX 40 - serial and replaced it with an AI good example .
DLSS 4 is still a assess thing for the GPU to run , but Nvidia is tackling the challenge with hardware improvements , too . It introduce fifth - gen Tensor cores , which are said to give up up to 2.5x the AI processing major power of the previous coevals . There ’s also hardware Flip Metering and an improved show locomotive .
Jacob Roach / Digital Trends
Overall , DLSS 4 will generate 15 out of every 16 picture element .
DLSS 4 will see adequate borrowing from the get - go . Around 75 game will put up DLSS 4 , and this include some of the AAA titles that will do good from it the most , such asIndiana Jones and the Great CircleandCyberpunk 2077 .
The Multi Frame Generation aspect of DLSS 4 is confine to RTX 50 - serial publication artwork card , but the residuum ofDLSS 4 is available on erstwhile RTX GPUs .
Image used with permission by copyright holder
DLSS 4: our impressions
Having covered what DLSS 4 really is , let ’s get into how it performs . In our reexamination of the RTX 5090 , our very own Jacob Roach put the card through its paces , let in DLSS 4 . And on newspaper , it ’s nothing unretentive of awesome .
As you could see in the above benchmark , DLSS 4 with 4x frame generation reach the unthinkable . The RTX 5090 , previouslystruggling to runCyberpunk 2077at max settingswith path trace enabled at 4 K , skyrocketed to a whopping 239 frame per 2d ( fps ) with DLSS Auto and 4x physical body genesis . It did that while cutting response time nearly in one-half , too .
Whether the RTX 5090shouldbe able-bodied to run an AAA game at max setting at above 30 fps is another account . What matters here is that DLSS 4 understandably occupy the biz from an unplayable state to something that can make the most of a 4k 240Hz monitor , and that ’s something — but search at visuals reveals some imperfections .
Image used with permission by copyright holder
While the frame rates remain satiny smooth , the plot itself does n’t experience that way . There are case of motion blur , artifacts , and objects that do n’t really take care natural .
Fortunately , Cyberpunk 2077might be an outlier , because DLSS 4 look best in other title of respect . Check it out above inMarvel Rivals . It seems that when the meanspirited frame charge per unit is at an satisfactory stratum , DLSS 4 does a solid job , even though it ’s not completely detached from artifact .
PlayingAlan Wake 2shows interchangeable results . Sure , there are some artifact , but none of them sense quite as unnatural as they did inCyberpunk 2077 — even though the fundament flesh charge per unit here also hovers at around 30 fps .
Image used with permission by copyright holder
DLSS 4 , while telling in its own right wing , may not quite be the showstopper that DLSS 3 was . Multi Frame Generation on its own does n’t always produce the results that DLSS 3 did , where you ’d just toggle it on and start playing . It ’s fantastic , but it might end up being more situational than the previous iteration .
RTX 50-series: performance
Nvidia made some important claim about the performance of these new 50 - series posting , saying they double the performance of their predecessors . As you could see in the sliding board above , that includes games , yes , but also AI and originative applications . The only direct compare Nvidia made was for the RTX 5070 , which it claims is now on par with the RTX 4090 .
That sounds like quite a bounce , but remember : it ’s done on the back of DLSS 4 ’s multi - frame generation feature . Most of Nvidia ’s carrying out claims and graphs include DLSS 4 , soit ’s difficult to experience exactly what the performance deltais when it comes to pure rendering .
Fortunately , with the RTX 5090 in the hands of reviewers , we can now put Nvidia ’s claims to the test with our own benchmarks of the RTX 5090 .
Image used with permission by copyright holder
It turns out that “ double as libertine ” was an exaggeration when it come to theRTX 5090 versus the RTX 4090 . Many expected outstanding carrying into action from the young Nvidia flagship , peculiarly pass on that theRTX 4090was roughly 80 % faster than the RTX 3090 . But , while the RTX 5090 is undoubtedly by far the libertine consumer GPU , it ’s definitely not doubly as fast as its herald .
Take a tone at our 4 K gaming benchmarks above , which showcase the middling Federal Protective Service based on a serial of thorough tests in 13 AAA titles . The RTX 5090 turned out to be around 30 % faster than the 4090 at 4 thou , but that ’s without Multi Frame Generation . We ’re speak about real frames here , and without DLSS 4 , the generational uplift is n’t quite as jaw - dropping .
That does n’t intend that the RTX 5090 isbad , however , and the norm is brought down by some games where the card break to deliver a meaningful improvement , such asAssassin ’s Creed Miragewith a mere 17 % cost increase . Meanwhile , inCyberpunk 2077 , the RTX 5090 well clear 100 fps and is around 54 % quicker . This is on ultra scene . We ’re view likewise significant profits inHorizon Zero Dawn RemasteredandDying Light 2 .
Image used with permission by copyright holder
Overall , there ’s no question : The RTX 5090isfaster , but it does n’t offer double the performance of the RTX 4090 — not even close .
scale down to 1440p reveals just how much you need a4 thou monitorfor a GPU like the RTX 5090 . There ’s no way to make the most of its likely withoutCPU bottleneck , and it designate , as the 5090 now only holds a 22 % star over the RTX 4090 .
It ’s extremely unlikely that anyone will apply an RTX 5090 with a 1080p admonisher , but we ’ve try out it just in case . Unsurprisingly , the constriction is real here , with just a 15 % jumper lead for the RTX 5090 .
Image used with permission by copyright holder
As far as shaft of light tracing goes , the RTX 5090 only serve to confirm the long - standing fact that Nvidia is king here . Sure , it ’s faster than the RTX 4090 , but no one expected anything else . This is yet another contemporaries where Nvidia is bound to rein the ray tracing game .
The RTX 5090 runsDying Light 2at 4 honey oil , with beam tracing and maximal preferences , at 90 fps — and that ’s without the Multi Frame Generation crutch provided by DLSS 4 . like results can be seen across every statute title we ’ve prove , aside from extremum likeCyberpunk 2077with way tracing enabled .
Overall , the RTX 5090 is a firm start to Nvidia ’s new generation of graphics cards , but it does n’t quite have the “ wow ” factor that the RTX 4090 came with . It ’s quicker , but not worth upgrading to from an RTX 4090 . We ’ll have to wait and see how the RTX 5080 fare to see whether it do up a solid generational jump .
Image used with permission by copyright holder
Should you upgrade to an RTX 50-series GPU?
At this microscope stage , with mountain of reviews to swear on , it ’s clear that the flagship GPUs in Nvidia ’s late line are top choices . But they might not beseem your budget at this stage .
Buying a new graphics card is seldom easy , even when you jazz on the nose what you want . OurGPU buying guidecan help ensure you make the correct selection for your microcomputer gaming apparatus and your budget .
